Aistudioworkspaceapproval MCP

What aistudioworkspaceapproval-mcp MCP server does

The aistudioworkspaceapproval-mcp MCP server connects an MCP-compatible client to a hosted AI Studio Workspace approval service. Its purpose is to place an approval step around workspace changes and model runs while producing structured records that can support review and auditing.

The service is hosted at https://aistudioworkspaceapproval.clauxel.com/mcp. It is a paid remote service rather than a package that runs from the repository. The repository contains public documentation and directory-submission material; it does not include the private production implementation.

How it works

An MCP client sends JSON-RPC requests to the hosted Streamable HTTP endpoint. Production requests must include a bearer token in the Authorization header:

Authorization: Bearer <token>

The token is obtained through the product website after selecting a plan and creating or requesting API access. The aistudioworkspaceapproval-mcp MCP server then exposes its approval, estimation, receipt, and audit functions through the configured remote connection.

Unauthenticated browser requests to the endpoint return a JSON error rather than internal service details. This endpoint behavior does not replace authentication for production tool calls; callers still need a valid token.

Setup and configuration

Setup requires a paid hosted-service account and an MCP client that supports Streamable HTTP remote servers.

1. Visit the product website and choose a plan.
2. Create or request an API token.
3. Add https://aistudioworkspaceapproval.clauxel.com/mcp as a remote MCP server in a compatible client.
4. Configure the client to send the token as a bearer authorization header.
5. Send JSON-RPC requests through the MCP connection.

No local server build or repository checkout is described as part of the setup. The public server card is available at https://aistudioworkspaceapproval.clauxel.com/server-card.json.

Tools and capabilities

The aistudioworkspaceapproval-mcp MCP server publishes five tools:

• check_workspace_change checks a proposed workspace change through the approval service.
• approve_model_run handles approval for a model run.
• estimate_ai_studio_spend estimates AI Studio spending.
• issue_workspace_receipt creates a structured workspace receipt.
• export_workspace_audit exports workspace audit information.

The supplied material identifies these tool names and their general roles, but does not document their input schemas, response formats, approval policies, pricing details, or receipt and audit field definitions. Consult the hosted service or its server card for implementation-specific details.

Limitations and notes

This is a paid remote MCP service, so availability and production use depend on access issued by the product. The README does not describe a local deployment option or provide private production source code. Client compatibility depends on support for Streamable HTTP remote MCP servers and bearer-token configuration; the documentation does not name specific MCP clients.

The service is described in terms of AI Studio Workspace approvals, structured receipts, audit logs, and reviewer-ready evidence. The provided material does not establish that it directly executes workspace changes or model runs, nor does it specify how a reviewer approves requests outside the exposed MCP tools.